On Mon, Jun 22, 2020 at 12:12:05AM +0530, gautamr...@gmail.com wrote:
> From: Gautam Ramakrishnan <gautamr...@gmail.com>
>
> Reduce image size of the image if all components have
> a non zero sample separation. This is to replicate the
> output of opj_decompress.
> ---
> libavcodec/jpeg2000dec.c | 19 ++++++++++++++-----
> 1 file changed, 14 insertions(+), 5 deletions(-)
>
> diff --git a/libavcodec/jpeg2000dec.c b/libavcodec/jpeg2000dec.c
> index ab36009a2d..05e85f4317 100644
> --- a/libavcodec/jpeg2000dec.c
> +++ b/libavcodec/jpeg2000dec.c
> @@ -269,6 +269,8 @@ static int get_siz(Jpeg2000DecoderContext *s)
> const enum AVPixelFormat *possible_fmts = NULL;
> int possible_fmts_nb = 0;
> int ret;
> + int o_dimx, o_dimy; //original image dimensions.
> + int dimx, dimy;
>
> if (bytestream2_get_bytes_left(&s->g) < 36) {
> av_log(s->avctx, AV_LOG_ERROR, "Insufficient space for SIZ\n");
> @@ -371,11 +373,18 @@ static int get_siz(Jpeg2000DecoderContext *s)
> }
>
> /* compute image size with reduction factor */
> - ret = ff_set_dimensions(s->avctx,
> - ff_jpeg2000_ceildivpow2(s->width - s->image_offset_x,
> - s->reduction_factor),
> - ff_jpeg2000_ceildivpow2(s->height - s->image_offset_y,
> - s->reduction_factor));
> + o_dimx = ff_jpeg2000_ceildivpow2(s->width - s->image_offset_x,
> + s->reduction_factor);
> + o_dimy = ff_jpeg2000_ceildivpow2(s->height - s->image_offset_y,
> + s->reduction_factor);
> + dimx = ff_jpeg2000_ceildiv(o_dimx, s->cdx[0]);
> + dimy = ff_jpeg2000_ceildiv(o_dimy, s->cdy[0]);
> + for (i = 1; i < s->ncomponents; i++) {
> + dimx = FFMAX(dimx, ff_jpeg2000_ceildiv(o_dimx, s->cdx[i]));
> + dimy = FFMAX(dimy, ff_jpeg2000_ceildiv(o_dimy, s->cdy[i]));
> + }
> +
> + ret = ff_set_dimensions(s->avctx, dimx, dimy);
> if (ret < 0)
> return ret;I think the resolution, the pixel format and any odd-format-handling code probably need to be adjusted together. Changing just the dimension code to select a higher resolution when chroma or alpha planes are of higher resolution then luma. Will probably not work if there is neither a pixel format selected that represents that nor code to compensate for this (as mjpeg does) but i might be missing something is there some testcase this fixes or improves ? thx [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB No great genius has ever existed without some touch of madness. -- Aristotle
signature.asc
Description: PGP signature
_______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
